Anchoring linguistic typology in universal semantic primes

In essence, "grammar is one and the same in all languages", but to establish what this universal grammar really looks like we have to investigate and compare many diverse languages, and for this we need a powerful and universally applicable metalanguage based on empirically established lexico-grammatical universals. The rough and incomplete outline of universal grammar sketched in this paper constitutes both a summary of the results arrived at by theoretical and empirical work over more than three decades (in the so-called "NSM" framework) and a program for further investigations. The author tries to show that it is possible to base investigations of universal grammar and typology on a truly universal, non-technical, non-arbitrary and intuitively intelligible tertium comparationis, and thus give it a secure and reliable foundation.